Dollar General

The Business Long Term

I grew up in the South with family in more rural parts but never shopped at DG. I knew my grandma shopped at Dollar Tree and equated the two in my mind. In 2020, a water line broke on my first floor, and I lived at my parent’s lake house for four months in deeply rural Georgia. The closest Walmart was 30 minutes away and the only option was a DG and a gas station. I wouldn’t have been interested in the company without that experience. It gave me a real sense of what “less competition” and “lower income” actually means and what “value plus convenience” means when a trip to Walmart would be 1.5 hours out of your day. I’ve read enough Buffett to know to stay away from retailers, and generally I think that’s wise. The ones he’s had success with like Nebraska Furniture and Borsheims have been the reinforcing effects of scale on demand (broader selection) and costs (economies of scale). It’s generally local market share that matters in retailing. That’s mostly what I saw when buying Carmax (now sold). It could sell more cars per location and have scale in production, distribution, advertising, etc. without ever being undercut on price. That combined with local market density (having many stores in one market) is a powerful combination. What it lacks was any sort of customer captivity.

DG has a local market share advantage but in a different way than is talked about. What matters is not DG’s total share of the market for things like health/beauty/food/perishables but its share of these type of purchases. Customers are still spending most of their money at grocery stores or Walmart. DG only captures 3-10% of HH spend with fill-in type purchases ($16 basket / 6 items) separate from a grocery trip. Only half of customers are shopping there monthly. For these type of purchases, DG dominates it’s local market. DG talks about their advantage as price plus convenience. Paying for convenience implies an active choice between price and time, and I can dial it back if I want to save money. That’s certainly the way I treat it. That’s a part of what DG does, but it misses a lot of what the company does for the very low-end consumer.

1. Convenience as small servings: DG customers don’t have the ability to make investments in bulk. There’s an anecdote from Cal Turner Jr’s book that their best seller in the 90s was $1 for 2 pack of popcorn. They tried to raise that to $2 for 8, quadrupling the value, and sales went down. Customer didn’t have the extra dollar to spend. DG is basically offering low absolute dollar, high dollar per volume products and trades higher margins for lower inventory turns. There’s a book called Nickle and Dimed where a reported tried to live on minimum wage. A theme that kept popping up was the inability to invest in the things that make sense long term (like buying in bulk) but are infeasible with limited savings. A lot of Walmart’s offerings, and clubs take this to the extreme, is passing savings through bulk offerings.

2. Convenience as liquidity: somewhat related, with limited savings, cash flows have to be managed much more aggressively around paydays, food stamps, and tax refunds for a sub $50k household. Being strapped for cash before payday mean not having the ability to buy more volume and not having the money for gas to drive 1 hour round trip to a Walmart. The money may be there next week for the big grocery trip, but today there’s a dollar left to spend on popcorn.

3. Convenience as time: Their shoppers skew female and both members of the family are working full time. Child care is an issue. This is the traditional way I think about convenience.

DG’s value proposition of pricing at parity with mass and providing convenience, broadly defined, is not something others are offering. Their moat is geography combined with scale (limited SKU purchasing scale, logistics, etc). There’s less competition in the small towns and the real competition is C-Stores and Drug Stores, rather than Walmart. Logistics are tough if you don’t already have it built. They’ve fortressed these small markets for the last ten years and the markets are going to grow slower than the overall US retail market, making a new entrant less likely. That incumbent scale combined with some consumer captivity based on of the convenience factor (customers are forced by their incomes to behave suboptimally) makes DG an attractive business in an otherwise unattractive industry. It’s also the reason I’m not interested in the Family Dollar turnaround. DG is 75% in small towns with less than 20k people. For Family Dollar it’s only 30%. The extra competition in suburban/urban markets makes that more difficult to predict how competitors will respond.

I think the above makes sense just rationally trying to think through the dynamics. But I keep coming back to the actual performance over the last 30 years which tell me their is a serious competitive moat here (the performance from 1967 IPO to 92 were just as good if not better while managed much poorer by the founding family).

1. 31 straight years of positive SSS prior to COVID; 34/36 positive since 1989. That’s a very weird stat given Walmart, Dollar Stores, Drug Stores, C-Stores are not a new phenomenon.

   1. Some of that is new stores adding 1-2% to comps; on $/SF basis it’s been positive 29/32 years (negative in 2000 -2%; 2006 -0.1%; 2017 -1.2).

2. Gross and EBIT margins have been more stable than any other retail concept since GFC.

3. Return on NTA (PP&E and NWC) has averaged 45% since 1992 (obviously there are leases as well)

   1. 40% in the 90s; 35% in the 00s; 60% in 10s

Of course naively extrapolating the past can’t be everything. Market and companies change. But when you see a long period of stability and customer behavior, it has to be weighted. What I currently see in the stock is near 100% weighting of the specifics of the recent past with very limited weighting of the base rate.

I start with these points to have more perspective on the longer term in the business being a good one if purchased at an attractive price. They have a moat that hasn’t changed because Walmart is investing in Ecom or comps have underperformed the last few quarters or inventory is (way) too high or management is (Maybe?) partly in denial or slow to respond on the issues.

---

The Stock

I’m buying the stock for the following reasons

1. I think it has structural competitive advantages that will last a long time

2. I think rural grocery stores are not going anywhere

3. I think you’re likely paying 10-11x run-rate earnings and almost certainly not above 15x

4. It’s priced as if it won’t grow, and I think it will

5. They have a consistent history of buybacks through many enviornments, which has the potential to add significantly to returns

6. I think you’re very likely to get at least 10% returns, and the probability of losing money over five years is low

Margins

Your returns in the stock are largely a question of what sustainable margins are. EBIT margins have average 7.8% since 1992, which includes 7.7% for the 90s, 6.5% for the the 2000s and 9.3% in the 2010s. They’ve been below 6% in five years total: 1991, 2000 (SEC settlement for accounting restatement that outsted Cal Turner Jr.), 2006-08 (just prior to KKR buyout where they restructured a lot of the business, large markdowns on packaway inventory, etc.). 80% of the time over the last thirty years, margins have been over 7%. Again, the past earnings record isn’t a perfect guide, but the inherent stability over very long periods of time coupled with some logic around the nature of competitive advantages leads me to weight it rather heavily versus what has happened in the last eighteen months. I don’t think the margins are going to be lower than 6% and I don’t think they’re going to be higher than 9%.

The argument for lower margins in the future are labor investment and inflation. Most of the investment they’ve made have been in hours versus wages. Some of the same dynamics around driving distances that holds for customers also holds for employees — driving one hour round trip can eat into more than half the incremental $2 you’d earn elsewhere. Getting labor has been difficult recently, and economic weakness can ease some of that. Not stretching GMs so thin with all of the things management has been trying to achieve the last few years can help as well. The increase in shrink is heavily correlated to GM turnover, so this is all connected at the store level. There were supply chain issues at the end of last year, mostly tied to the NCI initiatives, additional labor investment and generally poor execution. I believe these issues can be addressed. On the plus side they’ve been progressively taking out 25% markups through taking fleet in house and larger store formats come with higher margins.

I’d argue these are addressable rather than structural. The scarier element would be if you really thought their pricing was out of whack with mass or competition was increasing. These would be truly structural factors that the vast majority of retailers are subject to but I’m not overly worried about that with DG. They have excessive inventory right now. Markdowns are a certainity so I would expect margins to come down in the next 1-2 years.

At 6% and 9% EBIT margin, DG can make $7 to $11 EPS, which is 10x to 16x earnings. It’s very unlikely you’re paying too much for the business. I would lean more towards the upper end and say 8%, which allows for some margin compression vs. 9.2% over the last ten years. I think you’re paying about 11x normal earnings power for the business.

Growth

$PSF has grown at 2.7% over the last ten years. That includes share gains, which has mainly come from drug-store/convenience. Over the long-run this concept can’t grow as fast as suburban retailers for a couple of reasons. One, 75% of the footprint is in rural areas that are barely growing to slightly declining. Two, low wage workers haven’t until recently seen any growth in incomes. That may or may not be true going forward but they can’t price for long above customer incomes. Third, DG is way further ahead on optimizing this format then peers. Maximizing performance since the KKR buyout has been impressive and a major reason sales PSF are 30% higher than Family Dollar. This can’t be repeated. The lower growth in these geographies is effectively the “cost of carry” of the moat - lower growth but less competition (and less tempation to enter). In the long-run, there shouldn’t be a ton of real sales growth in this concept (with the corollary that it can’t trade at a big multiple either).

New units flatter SSS as they open at 80% of run-rate and ramp up. This adds 0.5%-2% to SSS depending on the rate of additions and is effectively double counting to add up SSS + new units. They’ve been opening 1,000 stores per year which has added 5% sales growth. They’re saying 12,000 total stores remaining, which would mean there is 13 years of growth left. There are reasons to be skeptical about that number. A lot of the new stores since 2016 have been fortressing markets as close as 8 miles from existing stores. Eventually cannibalization is going to be a bigger issue. The extensions of units into suburban/urban, POP concept, and talking about Mexico expansion is a hint that management is less confident in that 12,000. They’ve even made comments that the 12,000 is “available to everyone in the industry”, which is an interesting hedge. That said, returns are supposedly not falling off from 20% IRRs even burdening for cannibalization. Haircutting that number 40%, you’d get 8 years of remaining growth. I’m pretty confident there are at least 5 years, or 5,000 stores remaining. I wouldn’t count on more than 10 years of unit growth. At 11x earnings, the stock is priced like it isn’t going to grow at all.

An important risk here is that DG has a real estate machine that is going to be difficult to stop on a dime. When a management team has done the same thing for decades with incredible success and has a system that opens 2-3 stores a day, it would be very easy to ignore signs that returns are degrading quickly. This is combined with a comp structure that is heavily weighted to absolute profit dollars. While there is a ROIC component, management is clearly being paid to grow.

I’d give extremely high odds (say 9:1) to sales being higher in five years.

1. I think they face way less competition and a massive implosion in traffic is unlikely

2. The model is recession resistant mainly through trade-downs in the 50-70k income cohort. In every recession since 73/4, comps have accelerated through a downturn.

3. Even if SSS are weak, new unit growth does not take a lot of capital. They’ve been willing to grow through COVID, 08/09. They’re clearly incentivized to continue doing this and I think it’s unlikely the wall you’d hit on new stores is within the next five years

Yield/Capital Allocation

They can grow extremely efficiently. Growing HSD and building 1,000 stores annually has meant retaining only 25% of earnings — 15% for capex above D&A and 10% in working capital. If inflation persists, they’ll need to retain more to grow the same amount. That’s basically what happened in the last 18 months where FCF had to go towards inventory (and less funding from AP which I can’t fully understand). This should be normalized somewhat in the near term. Walmart moved quickly to markdown last years and Family Dollar a few quarters ago. The inventory situation is the most clear example to me of a management team in denial. The Non-Consumable Initiative has been underway prior to COVID, and I think they have had trouble reversing course and admitting they are way overbuilt on NC as mix shifted.

The historical growth pattern has been HSD Revenue/EBIT growth, 75% of Net Income converting to FCF, 30% getting paid out in dividends and the remainder in share repurchases for DD EPS growth. The basic risk to this formula is crowding out buybacks through:

1. More money goes into capex/inventory for the same growth rate - if inflation in steel/merchandise continues at a high rate there will be less available for real growth, Buffet’s “tapeworm” of inflation

2. Money gets diverted into debt paydown at very low returns. 3x lease adjusted leverage for a company selling everyday consumables doesn’t get me overly worked up. Maturities through 2026 can mostly be covered with cash flow if necessary, and covenant breakage implies 32% EBITDA declines / 5.1% EBIT margins. The bigger problem is shifting money into paying down the 1.5x turns of non-lease debt, which is much less attractive return wise than buybacks right now.

Buybacks are paused right now. They should have been paused late last year. They didn’t have the cash flow to do those buybacks and chose to add leverage to repurchase at high prices. The returns right now are very attractive. I think you can be a lot more confident repurchases will be resumed at DG than most companies. They’ve been doing it for decades in many different enviornments, so the institutional reflex is strong. Vasos has a long history of doing them, and board members are also aligned with it (Autozone CEO has been on the board since 09).

My expectation is repurchases are paused through most of 2024, FCF conversion is a bit worse at say 70% of net income, and the 30% dividend/ 70% repurchases continues thereafter.

Returns

Below is a simple five year forecast. It’s silly in it’s 20% below consensus in 24/25, which shows that all that really matters is your run-rate margin and exit multiple.

At the end of the day only a few things matter for returns to be 10% or better

1. Can they open new stores for at least five years (and comp sales not persistently negative)

2. Are EBIT margins going to be above 7%

3. Will it trade at least at 13x earnings

More importantly to me, to lose money you’d need to believe

1. They never open another store

2. EBIT margins are 6% or below

3. It trades at 11x or less

4. It’d lean towards mid-teens returns through a combination of 5% sales growth, 8% EBIT margins, 3% share reduction, and trading at 15x earnings.

Long at $109/share.